Vous êtes sur la page 1sur 10

American Economic Association

Decentralization Under Workers' Management: A Theoretical Appraisal

Author(s): Jaroslav Vanek
Reviewed work(s):
Source: The American Economic Review, Vol. 59, No. 5 (Dec., 1969), pp. 1006-1014
Published by: American Economic Association
Stable URL: http://www.jstor.org/stable/1810706 .
Accessed: 24/12/2011 21:17
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
American Economic Association is collaborating with JSTOR to digitize, preserve and extend access to The
American Economic Review.
Decentralization Under Workers' Management:
A Theoretical Appraisal
The purpose of my paper is to carry out
a theoretical appraisal of a labor-managed
economy of the Yugoslav type. In my opin-
ion, it is this type of economy that repre-
sents, by and large, the true aspirations of
reformers in eastern Europe.
There is full justification for such an en-
deavor. On the one hand, the empirical
studies which we have on the subject are not
fully conclusive; on the other, even if quanti-
tative results fulfilled some accepted criteria
of statistical significance, we still might want
to verify them through a theoretical evalua-
tion. Moreover, it ought not to be forgotten
that the Yugoslav experiment is unique and
comparatively young. To compare its real
performance with that of systems which
have been tried in a large number of in-
stances over long periods, is not quite fair.
At least as a supplement to the emipirical
studies, it is thus necessary to compare the
corresponding theoretical models.
Besides the direct and practical objective
of my paper, related to the economies of
eastern Europe, there is also the more ac-
ademic question of the theoretical literature
on the subject at hand. The latter, although
of excellent quality, is extremely limited. In
fact, there are only two important articles
on what I would call economics of labor
participation-those by Benjamin Ward
[7] and
Evsey Domar
[3]. These authors do
not make anything near a full evaluation of
an entire economic system. And yet, the
findings emerging from such a limited cover-
age often are taken as characteristic of the
efficiency of the system. For example, Ward's
perverse supply elasticity in the short run is
taken by many as the proof of the absurdity
of labor management. Perhaps the best and
most authoritative illustration of the overall
pessimism regarding the labor-managed econ-
omy is contained in Abram Bergson's more
recent evaluation of market socialism [2].
At this stage of the argument, I do not
want to dispute any specific points regarding
the labor-managed economy; that will be
done explicitly or by implication later in our
discussion. I only want to contend that to
appraise an economic system it is necessary
to consider all of its major aspects. And this
is what I propose to do.
The task would be an impossible one if
each point were to be fully explained or
proven. Fortunately, I have produced a more
complete analysis elsewhere [5], [6], and
thus I can restrict myself here to the pre-
sentation of the main conclusions, supple-
menting these with brief indications of the
underlying reasoning. I will further limit
my analysis by concentrating on questions
of global economic efficiency, leaving out, as
much as this is possible, the mechanics of the
I will first consider what may be referred
to as a dehumanized model where, as in
conventional capitalist theory, labor is con-
sidered merely a factor of production, of
constant quality, exogenous to the system.
This is done in Sections I and II In Section
III, by contrast, I discuss some of the most
important special dimensions which emerge
from the participatory nature of labor man-
agement. The principal objective of the re-
maining section is an overall evaluation of
the labor-managed system in comparison
with other major world systems.
I. The Pure Model
In what I call the pure model, I make basi-
cally the same assumptions as those under-
lying the contributions of Ward and Domar
noted above. They can be summed up as
implying a perfect, competitive and smooth
neoclassical world in which the moving force,
contrary to the capitalist situation, is niaxi-
Cornell University.
mization of income per laborer. There is only
of labor, perfectly homogeneous,
and active only as a factor of production.
The only characteristic emanating from
labor management is income sharing and the
behavioral principle of maximization just
noted. It should be clear from the outset that
an economy thus defined is f ar from the
complex reality of Yugoslavia, or any other
econiomy adopting labor management; how-
ever, in my opinion the assumptions capture
the ideal form of the economy, and thus
their implications should be studied as a
first step in any comprehensive evaluation.
To deemphasize the single firm approach
used by WVard and Domar, it may be desir-
able to reverse the process, and start with
the discussion of a complete full employment
general equilibrium solution of the labor-
managed economy (situations involving un-
employment will be taken up in the next
Let our starting point be the ideal condi-
tions of perfect competition and full em-
ployment. We may make the observation
that when all firms of an industry use the
same technology and free entry is guaran-
teed, the labor-managed economy will be
Pareto optimal. In other words, just like its
ideal capitalistic counterpart, the labor-
managed economy will be producing the
maximum producible output from given re-
sources, and the maximum social satisfaction
for a prescribed distribution of income.' These
conclusions follow from the fact that com-
petitive labor-managed firms equalize factor
marginal products to factor returns for all
factors including labor, from coinpetition in
nonlabor factor markets, free entry of firms
and identical technologies.
Bv contrast, as has been pointed out by
Ward [7], if technologies of different firms
within an industry are different, the opti-
mum solution will not be reached by the
labor-managed economy because the equilib-
rium behavior will not lead to equalization
of marginal value products of labor among
firms. In this context the capitalist alterna-
tive appears as superior, on the assump-
tions made. Of course, if imperfections in the
labor market are permitted, the differences
in marginal productivities of labor among
firms and industries may be just as impor-
tant. In any case, the comparative ineffi-
ciency of the labor-managed case here dis-
cussed should not be too pronounced. In
fact, the labor-managed firms are all bound
to operate at a point of maximtum factor
productivities (i.e., where their technologies
are linear-homogeneous), the comparative
shortcoming being merely imputable to the
fact that the more efficient ones do not pro-
duce enough. Moreover, as we argue later in
Section III, labor-managed economies will
normally be in a much better positioIn to
proliferate the best technology throughout
the industry. Finally, even if in the case
studied the labor-managed economy will pro-
duce less than the capitalist with identical
factor endowments, it can be shown that it
will generate a stronger demand for, and
correspondingly higher returns to, capital.
Thus accumulation and growth may pro-
ceed at a higher rate, leading to more output
and more consumption over time.
While recognizing the abstraction and
lack of realism of all of the pure theory pre-
sented in this section, it may be said that the
perfectly competitive model just discussed
is the least realistic. Indeed, the require-
mients of product homogeneity and a very
large number of sellers (producers) are satis-
fied only in a few industries. Much more
frequent is some degree of monopoly power
coupled with product differentiation and, in
many instances, active sales promotion. It is
in this context that, I believe, Ward's earlier
conclusion of greater restrictiveness of labor-
managed monopolies, while formally correct,
is misleading when we want to evaluate the
theoretical performance of the system.
It is true that if the government or some
other external agent were to set the number
of firms for each industry at one, such a
monopoly would be, ceteris paribus, more
restrictive and thus socially more harmful
than a capitalist monopoly. But in the real
world monopolistic tendencies and market
power are hardly ever of this type. Rather
We will speak of the distributional effect.s of labor
nanagement in Section T11.
they derive from the fact that efficient pro-
duction is consistent with only a limited
number of firms given the size of the market,
the desire of firms to accumulate and grow,
the artificial creation of barriers to entry,
If viewed in this more realistic context of
the entire
industry, one can make a case for
labor management which is strong indeed.
On grounds of several arguments, labor
management can be expected not only to
Xyield market structures more competitive
than any other free economy, but also pre-
vent a good deal of wasteful and harmful
sales promotion. The first and perhaps sim-
plest argument is that with an increasing
scale of operation, the benefits from partici-
pation-incentives, identification, and in-
volvement-will tend to diminish.
Thus it can be expected that, all other
things being equal, the point of maximum
efficiency, which also is the point of long-run
equilibrium for a labor-managed firm, will
be reached for a lower level of output than
for a firm operating with a hired labor factor.
And thus, figuratively speaking, there will
be room for more firms in a given industry.
The next point also hinges on the most
efficient scale of operation. The labor-man-
aged firm will never grow beyond that scale,
whereas a capitalist firm often will; its growth
being governed, even after greatest technical
efficiency is reached, by the desire for profit
maximization. Two extreme situations may
further clarify the essence of this argument.
Consider an ideal firm operating under con-
stant returns to scale and facing a constant
price yielding positive unit profits under
capitalism. The equilibrium level of opera-
tion of that firm, if capitalist, is (at least in
theory) infinite, while if labor-managed, it
is finite and indeterminate at any level of
operation. The simple crux of this deduction
is that while the first firm must grow in-
definitely to maximize profits, the second
maximizes income per laborer at any level
of operation. The other extreme situation is
similar to that just described except that the
firms in question face a less than infinitely
elastic demand function. In that situation
the capitalist monopoly, as is well-known,
will find its equilibrium where the marginal
(and average) cost reaches the marginal
revenue; whereas the labor-managed firm
must operate at zero output, or, more realis-
tically, with only one employee; indeed, as
the reader will easily verify, under the as-
sumed conditions income per laborer will be
maximized with only one member of the
labor-managed firm. Clearly, under the as-
sumed conditions it would take a very large
number of firms to fill the industry (with a
reasonable return to labor), but by the time
such a large number would enter, the market
power of each firm would become very low or
disappear altogether.
Probably the most significant comparative
advantage of the labor-managed oligopoly
arises in connection with product differen-
tiation and sales promotion. It can be shown
that the labor-nianaged firm will in equi-
librium produce less and engage in a less
intensive advertising campaign than an equiv-
alent capitalist firm as long as the latter
makes positive profits, a condition which
obtains virtually without exception. Espe-
cially if we realize that it is precisely the more
extreme doses of exhortational advertising
that constitute the heaviest social costs, the
advantage of the labor-managed alternative
is considerable. Not only can we expect a
significantly less concentrated industry, but
the external diseconomies, mind pollution,
etc., should be significantly reduced or com-
pletely eliminated.
The last point deserving mention is based
on the simple notion that there is a far
greater desire within democratic structures
to decentralize than in nondemocratic ones.
For example, even with very small or no
economies of scale, a capitalist, or for that
matter a Soviet-type, firm will tend to retain
its centralized organization, whereas the
labor-managed one will attempt to subdivide
itself into autonomous decision making and
production units based on location or other
functional characteristics. I am quite con-
vinced that this tendency is present with the
labor-managed alternative for other reasons
than the stronger production-incentive of
small groups.
Finally, before concluding this section, we
ought to reconsider Ward's negative supply
elasticity, and discuss that and related is-
sues in the context of general equilibrium
efficiency. It is true that if short-run reactions
of competitive firms were as posited by Ward,
the resulting inefficiency could be quite con-
siderable, and what is equally important, at
least some markets could become unstable.
The contention which I want to argue below
is that matters are far from being so bad.
First of all, we have the two arguments
made bv Domar: One,
with more than one
product by the firm, even short-run elasti-
cities can be positive and quite high for an
individual product; two,
even with only
one product the supply elasticity in the short
run must be positive if the firm operates
with an (active) external labor supply con-
straint, reflected by a labor supply function
of finite elasticity [3].
In addition, there is
the problem raised by Joan Robinson [4] in
her critique of Domar. Using my own words
to push the argument a step further, how
can one reasonably expect that a working
collective will mutilate itself (kicking out
say 1/10 of the membership), if it has al-
ready realized a significant gain from a price
increase, say 10 percent of income, for the
sake of gaining an extra, say, one percent?
Indeed, this sounds like an extract from a
book of rules of capitalist conduct. While
what underlies the above rhetorical question
certainly is true, I would further like to
point out that the elimination of the nega-
tive elasticity argument does not even call
for any higher morals. In the
single-product producer operating under
perfect competition will have very little
possibility of capital-labor substitution; in
terms of Ward's diagrams, the relevant
stretch of the marginal productivity of labor
will generally be very near a vertical line.
And, consequently, the normal short-run
supply elasticity of a competitive labor-
managed firm producing a single output can
for all practical purposes be considered zero,
even if group solidarity were absent.
Before leaving the short run, two obser-
vations are in order. First, let it be noted
that the zero elasticity eliminates possibili-
ties of instability, and reduces a good deal
the loss of general equilibrium efficiency
(note that in a world with only single-prod-
uct firms a change in relative prices would
now keep, in the short run, the point on the
production possibility locus unchanged). Sec-
ond, in the context of aggregate national
income analysis, the zero short-run supply
elasticity by no means is something to frown
on; but this we will discuss in the next sec-
Finally, a few words on long-run adjust-
ments within the general equilibrium frame-
work. Again, the introduction of an addi-
tional piece of information from the real
world helps a good deal. It is reasonable to
expect that with most firms the point of
constant returns to scale (the point corre-
sponding to the minimum of the long-run
average cost curve) in reality is not a point,
but a whole range (of "efficient outputs")-
a range the end of which is hard to establish
empirically. But once in such a range, as we
have noted already, the long-run equilibrium
of the labor-managed firms becomes indeter-
minate. The firms now can arbitrarily de-
termine their scale of operation, and the
most logical is that given by the size of the
working collective, more or less exogenously
determined. But as the reader will find easy
to verify, under these conditions the long-
run supply elasticities will be positive and
can be quite high even with single-product
firms, and the general equilibrium adjust-
ment to changing demiand conditions will be
quite efficient indeed. Of course, the most
efficient and full (Pareto) optimal adjust-
ment will occur once labor incomes are
equalized through entry and exit-or, with
firms operating under constant returns to
scale, through expansion or contraction of
existing firms.
II. Macroeconomics
We now turn to the determination of ag-
gregate variables such as income, employ-
ment, the price level, etc. I feel that it will
be most expedient first to summarize the
principal results that can be obtained for the
labor-managed economy and then, to the
extent that space permits, to elaborate on
some of them.
The first observation is that unemploy-
ment is conceivable in a labor-managed econ-
omy, but if it occurs it will be of an entirely
different nature--because of different causes
and different duration-than unemployment
of the Keynesian type in a capitalist econo-
my. A second imiportant point is that varia-
tions in effective demand can be expected in
the labor-mianaged economy to lead pri-
mnarily to variations in prices and not much
or not at all to adjustments in income and
employment. A third and related point is
that on the whole there will be little natural
cause for secular inflation, although varia-
tions in prices may be wide if markets are en-
tirely free. Fourth, even if unemployment
arises there will always be natural forces to
restore full employment after a while, pro-
vided that capital markets are competitive.
Fifth, the macroeconomic general equilib-
rium determining simultaneously the aggre-
gate variables must be stable provided that
real cost of capital varies with the general
price level, and almost certainly it will be
stable even if the money cost of capital
is constant.
The key to points one and two is the low or
zero short-run supply elasticity of firms. If
for one reason or another the existing number
of firms at given product and factor prices is
incapable of employing (in equilibrium) all
the labor force, then unemployment will pre-
vail, and the supply elasticity being low,
short-runi changes in demand will be trans-
lated into price level variations. Note that
this has nothing to do with wage-rigidity.
Such rigidity in fact is highly unlikely if not
impossible in the labor-managed economy
because of the residual nature of labor in-
come and the direct managerial ability of
labor to make choices between price reduc-
tions on the one hand and idleness and sales
reductions on the other. Thus, contrary to a
capitalistic situation, a symmetrical flexi-
bilitv of prices, up and down, and no syste-
mlatic inflationary tendency should be ex-
Point four is a logical consequence of the
fact that if there is unemployment, the un-
employed will be in a considerably favored
position in comipeting with existing firms for
capital funds for development of new proj-
ects. Indeed, with very low unemployment
incomes those currently unemployed will be
able to offer considerably better terms to the
lender. And thus any significant degree of
uniemployment carries with itself strong,
even if perhaps not overly speedy, self-
correcting forces. The prime vehicle of the
latter is entry, and perhaps also expansion
of existing firms (often into new lines of pro-
duction) operating under constant returns to
scale, whose scale of operation is arbitrary.
It should be noted here that no parallel rem-
edy exists for Keynesian unemployment;
here either entry or expansion of existing
firms implies opening of a deflationary gap
and the ensuing return to a less than full em-
ployment equilibrium.
Point five asserting, stability of the inacro-
economic system involves some slightly in-
volved mathematics and consequently I
omit its demonstration. Full proof is con-
tained, however, in my General Theory of
Labor-Managed Market Economies [5].
We may suTn up by taking the position of
the policy maker in the labor-managed econ-
omy: He need not fear autonomous varia-
tions in demand as a cause of cyclical un-
employment. Rather he may look out for
such variations in order to prevent fluctua-
tions in prices. In this endeavor he will be
greatly assisted by an open and competitive
foreian trade sector. Full employment does
not entirely disappear, however, as the policy
maker's problem. But he must be concerned
with it as a matter of long range strategy, pri-
marily ensuring perfect functioning of the
capital market with proper interest struc-
tures. More actively, he may want to take
discretionary steps: market research and
consulting, indicative planning, forecasting,
etc., to enhance the speed and efficiency of
new entry and expansion. In the final analy-
sis, however, he has a considerable advantage
compared to his capitalist colleague in that
matters are bound to improve even if per-
chance be fell asleep on the job.
III. Special Dimensions of Labor Management
Even if the labor-managed economy can
pass with flying colors the scrutiny of con-
ventional micro- and macro-economic theory,
as presented in the preceding two
its greatest strength lies in what we
identify as its special dimensions; diinensions
larrgely- absent in other economic systemns.
Thev are all related to, contained in, or emer-
gent from the managerial function of labor,
that is, of all participants of the enterprise.
The field of analysis opened bv these special
dimeinsions is so vast that it cannot be
treated inl a single section; consequently we
will restrict ourselves here to the outliine of a
few of the most important arguments. In my
larger studv referred to earlier, I devote to
the subject several chapters [5, Part III],
and even there, I certainly cannot pretend
that my analysis is exhaustive.
The first and simiplest consideration is
that, contrarv to the models studied in the
two preceding sections, labor is not unique
and homogeneous, but rather that in every
enterprise a large nuumber of individuals of
different skills and qualifications cooperate
in a comnmon endeavor. One necessary task of
labor management thus is to decide on the
distributive shares among the different labor
categories. The specific forml of the distribu-
tion schedule, indicating whether the director
gets four times the pa.y of the janitor, etc.,
will be the outcome of two sets of forces:
Conditions of the labor market or mnore pre-
ciselv, of the quasi-labor market, because
there is no conventional labor market in a
labor-managed economv; and the collective
will of the working com-nmunity as it emerges
from the democratic decision making process
of labor management.
Noting that the labor-manaeged firm in
equilibrium (i.e. when maximizing income for
each laborer) will equalize the income of each
labor categorv to the corresponding mnar-
ginal value product, the first set of forces
will guarantee that the allocation of labor
throughout the economy will be at least ap-
proximately consistent with maximum ag-
gregate output.2 The second set of forces, on
the other hand, will guarantee a reasonable
distribution of income, consistent with the
generally accepted notion of justice for a par-
ticular segment of
a particular com-
munity, or a particular period of time. This
"dual" optimization of social welfare, strik-
ing a balance between the mechanistic rules
of efficient resource allocation and a collec-
tive expression of distributional justice,
seems to be a solution superior to that offered
by either set of criteria in isolation. The car-
dinally important thing is that the "dual"
system guarantees that major mistakes will
not be committed in either the allocational or
the distributional sphere.
The labor force in a labor-managed econ-
omy is not only diverse but also of variable
quality. And this brings us to the second im-
portant special dimension of labor manage-
ment. Without any- doubt, labor-manage-
ment is
all the existing forms of enter-
prise organizations the optimal arrangement
when it comes to the finding of the utility-
maximizing effort, i.e. the proper quality,
duration and intensity of work, by the work-
ing collective. Not onlv is there no situation
of coInflict between management and the
workers that might hinder the finding of the
optimum, but the process of self-manage-
ment itself can be viewed as a highly efficient
device for comnmunication, collusion control
and enforcement among the participants.
This should be contrasted with a situation
of most other einterprises where the worker
normally will be furnishing a minimum effort
consistent with the retention of his job. Is is
true that expectation of promotion or a raise
mnay stimulate effort over the acceptable
minimum, but recall that this factor should
be equally operative in all firms, and thus it
does not establish a comparative disadvan-
tage of the labor-managed firnm.
Our second special dimension may con-
veniently be summarized in terms of a dia-
gram. In Figure 1 we measure effort (E) to
the left of 0 along the horizontal axis,
whereas on the vertical axis we find a (aver-
age) laborer's income Y. A typical transfor-
mation function for the labor-managed alter-
native between effort and income is indi-
cated by the contour ab (note that with some
fixed costs it takes some effort to generate
In pure theory a frictionless system should genierate
perfect income equalization through entry and/or exit
of firms. However, considering that in the real world
there is never enough time to bring such an equalization
process to completion, given imnperfect information,
attachments and loyalties of individuals to the working
collective, etc., the competitive forces can be expected to
prodiuce only aPProximate results.
(E) -\
zero income). Bv contrast, a typical trans-
formation function in capitalist enterprise is
indicated bv the rectangular broken contour
defined by point k. Its comiparative defi-
ciency is of two kiinds: 1) Point k is below ab,
on account of some residual profit normally
not distributed amnong the members, and 2)
and much more important, ab is continuous.
On account of both, (as the reader may verify
by drawing a set of convex indifference lines)
losses in individual and collective utility,
that is, in real income, will occur in one type
of undertaking as coimpared with the other.
Third, as shown by the diagram, there is
lanother special dimension, Z-in fact, it
should be understood as a summlary index
standing for a very large numiiber of special
dimensions--also entering together with in-
come and effort, the utility function of the
members of the working collective. As irndi-
cated by the three dimensional transformia-
tion locus in Figure 1, attainmeilt of higher
levels of Z involves an opportunity cost in
terms of income and/or effort. Nevertheless,
this new dimension further augments the
number of choices of the working community
and thus unambiguously (as shown by the
equal-utility contours u, u', u" and the
maximum-utility point e) its real income. It
is the comparison between point e and point
k that should, in the context of our present
analysis, be included in any comparison be-
tween a labor-managed and a capitalist, or
for that matter, Soviet-type firm.
It would hardly be possible to be exhaus-
tive in explaining all the variables for which
Z actually stands. However, some illustra-
tions should be helpful: For example, the
working collective, or a subgroup of it, may
prefer a conveyor belt to a traditional way of
intraplant transportation, even if this should
involve a reduction in money income. By
contrast, a capitalist would perform such a
substitution only if profit could be increased.
Another example is training and education of
some members of the community at the ex-
pense of the enterprise. Again, the corre-
sponding motivation and decision making
is entirely different from what it would be in
other firms. Still another example is collec-
tive consumption using part of the global
income of the enterprise, including housing,
recreation, etc. Moreover, the enterprise as a
whole may undertake at its own expense
social action directed toward the outside, the
local community or other, and thus derive
intangible real income-we may call it peace
of mind-for itself.3
The last example brings us to a fourth
category of special dimensions, related to Z
but not exactly the same, referred to as Z'.
It resembles Z in all respects, including that
it enters the utility index, except that it does
not affect money income or effort of the
working community. The transformation
locus with Z' substituted for Z would thus
appear as a cylindrical surface generated by
lines parallel to the Z' axis. Normally, such a
suiface will be truncated at some finite level
of Z' and, obviously, an equilibrium will be
found (with indifference loci convex every-
where) at the corresponding ridge, at a level
of collective utility again unambiguously
higher than at a point such as d with two di-
mensions only. To give substance to this
rather technical exposition, let us give at least
Some will object that all this can be done by individ-
uals or through nonbusiness-based groupings: Tbis is
true, but it holds even for the labor-managed system and
thus the potential of action by firms, the primary gen-
erators of resources, still constitutes a comparative ad-
one example; it is a significant one. We have
noted in Section I that the long-run equilib-
rium of a competitive labor-managed firm
operating under constant returns to scale-a
condition encountered in many real situa-
tions-is indeterminate with respect to scale
of operation. Under such conditions, total
employment and output of the firm may be-
come our Z' variable. The working commun-
ity considers it a positive utility to maximize
employment in the community; the external
constraint, and the ridge line of our above
discussion now are given by the fuill emplov -
ment in the community.
In the opinion of the writer, the argument
just made has a good deal of relevance, and
when further generalized can become even
more significant especially in the context of
less developed countries. The generalization
is based on the recognition that in most situa-
tions-especiallv in manufacturing- e can-
not postulate perfect competition; some
market power will always be present. In
such situations, however, all it takes is to re-
classify Z', a zero opportunity cost variable
into Z, a positive opportunity cost variable.
But the absorption of unemployment in the
community remains a positive good from the
point of view of the working collective. Thus
a solution involving higher employment than
that dictated by strict income-per-laborer
maximization still may be expected.
IV. Concluding Observatitons
Before attempting a summary evaluation
of the labor-managed economy, two more
matters ought to be taken up which norm-
ally enter the appraisal of an economic sys-
tem. The first is the question of ability to
mobilize national savings for accumulation.
The second is the question of the capacity of
generating technical progress and innova-
Regarding the capacity to mobilize invest-
ment funds, the labor-managed economy,
especially its socialist version, is in a very
strong position. Noting that the labor man-
aged firms must pay for the use of capital
the corresponding value of marginal product,
and taking a reasonable set of estimates of
national capital stock and marginal capital
productivity, the economy can generate net
savings of between 20 and 30 percent of na-
tional income just by retaining capital in-
come for reinvestment. These percentages,
already among the highest currently en-
countered in the real world, can further be
augmented through private savings, govern-
ment surpluses, or as in
an obligation of firms to repay investment
loans from current income.
With respect to the second question,
case for the labor managed economy is some-
what less clear cut; but in the final analysis,
it would be difficult to conclude that the sys-
tem in question is at a comparative disad-
vantage. When it comes to "small" inventive
and innovative activity emerging as a side-
effect of the productive process itself, the
incentives, unity of purpose and ease of
communication offered by labor management
(and by income sharing, which is an integral
part of labor-management) are conditions
unequalled by any other system. On the
other side of the spectrum, with respect to
basic research and major scientific develop-
ments, the firms only rarely play a decisive
role; thus in the present context the differ-
ence between alternative systems cannot be
very important.
It is the middle of the spectrum, for "me-
dium size" incentive and innovative activity,
that would seem the most important. Here
the modern capitalist corporation almost
certainly has a comparative advantage with
its own low-cost-and often quite abundant
--funds that it can employ in research and
development. On the other hand, a socialist
economy, labor-managed or other, is likely
to have a distinct advantage in proliferating
innovations throughout the whole economy.
Let us now try to make a sumnmary evalua-
tion of the labor-managed system. I realize
that such an endeavor must always contain a
certain element of subjective judgment. This
is so because summary evaluation really
rimeans constructing an aggregate ordinal in-
dex based on a combination of several evalu-
ations, not all of the same sign; clearly, some
weights must be employed and these, to a
degree at least, will be subjective.
With this in mind, and basing myself on
ten years of inten-sive study more than on the
present exposition, I cannot forego a set of
strongly favorable conclusions; I cannot
avoid them, even though I realize that I am
contradicting the majority of our profession
who have thought about the problem, and
even though I may risk earning the displea-
sure of many.
In brief, the labor-managed system ap-
pears to me to be superior bv
far, judged on
strictly economic criteria, to any other eco-
nomic system in existence. In the sphere of
allocation efficiency (concerning how well it
utilizes its resources in inational production),
it is at worst equal to the western-type cap-
italist system in the context of a full-em-
ployment model (as discussed in Section I),
while it is definitely superior in the context of
the macro-economic model (as discussed in
Section II) and in the context of its special
dimensions (as discussed in Section III). On
the side of the system's capability to grind
out an efficient pattern of income distribu-
tion, there are also strong reasons to believe
that a socialist labor-managed economy will
do a better job than other market systems.
This conclusion is based not onlv on the argu-
ment of collective decision miaking within the
firm (as argued in Section III), but also on
the fact that in the socialist econonmy the
income share of capital, whether reinvested
or not, will accrue to the society as a whole
and not to a select group of individuals.
Compared to the Soviet-type command
economy, the question of distributional
efficiency really is empty of meaning because
income distribution is decided by decree and
not by some mechanism inherent in the
economic system. As for allocational effi-
ciency, far more important for our compari-
son because inherent in the system, we only
have to recognize with Bergson [1] that the
Soviet-type command economy is less effi-
cient thani market capitalism, and recall the
above evaluation of the two market alterna-
tives. In fact, even the weaker postulate of
approximate equality between the two major
world systems would suffice here to establish
a preference for the labor-managed alterna-
tive over that of a command economy.
Two remarks less strictly economic, and
stemming much more from intuition than
from careful analysis, may be in order before
closing this discussion: First, it seems to me
that the comparative advantage of labor-
managed systems becomes even stronger
once we leave the strictly economic frame of
reference and replace it by one that takes
account of broader human values. Second,
taking a very long view of world events, it
seems to me that if there is a meeting ground
for the presently conflicting major world sys-
tems and ideologies, it is one not too far from
the system discussed here.
1. A. BERGSON, The Economics of Soviet
Planning, New Haven, 1964, p. 341.
2. "Market Socialism Revisited,"
J. Polit. Econ., Oct. 1967, 75, 432-42.
3. E. DOMAR: "The Soviet Collective Farm
as a Producer Cooperativ e" Amer. Econ.
Rev., Sept. 1966, 56, 734-57.
4. J. ROBINSON: "The Soviet Cooperative
Farm as a Producer Cooperative: Com-
ment" Amer. Econ. Rev., Mar. 1967,
5. J. VANEK, The General Theory of Labor-
Managed Market Economies: to be pub-
6. , The Participatory Economy:
Evolutionary Hypothesis and a Develop-
ment Strategy, to be published.
7. B. WARD, "The Firm in Illyria: Market
Syndicalism" Amer. Econ. Rev., Sept.
1953, 58, 566-89.